Add some instrumentation to the bgwriter, through the stats collector.

New view pg_stat_bgwriter, and the functions required to build it.

Author: mhagander

doc/src/sgml/monitoring.sgml

@@ -1,4 +1,4 @@
-<!-- $PostgreSQL: pgsql/doc/src/sgml/monitoring.sgml,v 1.47 2007/03/16 17:57:35 mha Exp $ -->
+<!-- $PostgreSQL: pgsql/doc/src/sgml/monitoring.sgml,v 1.48 2007/03/30 18:34:55 mha Exp $ -->
 
 <chapter id="monitoring">
  <title>Monitoring Database Activity</title>
@@ -260,6 +260,16 @@ postgres: <replaceable>user</> <replaceable>database</> <replaceable>host</> <re
      </entry>
      </row>
 
+     <row>
+      <entry><structname>pg_stat_bgwriter</></entry>
+      <entry>One row only, showing cluster-wide statistics from the
+      background writer: number of scheduled checkpoints, requested
+      checkpoints, buffers written by checkpoints, lru-scans and all-scans,
+      and the number of times the bgwriter aborted a round because it had
+      written too many buffers during lru-scans and all-scans.
+     </entry>
+     </row>
+
      <row>
       <entry><structname>pg_stat_database</></entry>
       <entry>One row per database, showing database OID, database name,
@@ -751,6 +761,71 @@ postgres: <replaceable>user</> <replaceable>database</> <replaceable>host</> <re
       </entry>
      </row>
 
+     <row>
+      <entry><literal><function>pg_stat_get_bgwriter_timed_checkpoints</function>()</literal></entry>
+       <entry><type>bigint</type></entry>
+       <entry>
+        The number of times the bgwriter has started timed checkpoints
+        (because the <varname>checkpoint_timeout</varname> time has expired).
+       </entry>
+     </row>
+
+     <row>
+      <entry><literal><function>pg_stat_get_bgwriter_requested_checkpoints</function>()</literal></entry>
+      <entry><type>bigint</type></entry>
+      <entry>
+       The number of times the bgwriter has started checkpoints based on
+       requests from backends because the <varname>checkpoint_segments</varname>
+       has been exceeded or because the CHECKPOINT command has been issued.
+      </entry>
+     </row>
+
+     <row>
+      <entry><literal><function>pg_stat_get_bgwriter_buf_written_checkpoints</function>()</literal></entry>
+      <entry><type>bigint</type></entry>
+      <entry>
+       The number of buffers written by the bgwriter during checkpoints.
+      </entry>
+     </row>
+
+     <row>
+      <entry><literal><function>pg_stat_get_bgwriter_buf_written_lru</function>()</literal></entry>
+      <entry><type>bigint</type></entry>
+      <entry>
+       The number of buffers written by the bgwriter when performing a
+       LRU scan of the buffer cache.
+      </entry>
+     </row>
+
+     <row>
+      <entry><literal><function>pg_stat_get_bgwriter_buf_written_all</function>()</literal></entry>
+      <entry><type>bigint</type></entry>
+      <entry>
+       The number of buffers written by the bgwriter when performing a
+       scan of all the buffer cache.
+      </entry>
+     </row>
+
+     <row>
+      <entry><literal><function>pg_stat_get_bgwriter_maxwritten_lru</function>()</literal></entry>
+      <entry><type>bigint</type></entry>
+      <entry>
+       The number of times the bgwriter has stopped its LRU round because
+       it has written more buffers than specified in the
+       <varname>bgwriter_lru_maxpages</varname> parameter.
+      </entry>
+     </row>
+
+     <row>
+      <entry><literal><function>pg_stat_get_bgwriter_maxwritten_all</function>()</literal></entry>
+      <entry><type>bigint</type></entry>
+      <entry>
+       The number of times the bgwriter has stopped its all-buffer round
+       because it has written more buffers than specified in the
+       <varname>bgwriter_all_maxpages</varname> parameter.
+      </entry>
+     </row>
+
      <row>
       <entry><literal><function>pg_stat_clear_snapshot</function>()</literal></entry>
       <entry><type>void</type></entry>

src/backend/catalog/system_views.sql

@@ -3,7 +3,7 @@
  *
  * Copyright (c) 1996-2007, PostgreSQL Global Development Group
  *
- * $PostgreSQL: pgsql/src/backend/catalog/system_views.sql,v 1.36 2007/03/16 17:57:36 mha Exp $
+ * $PostgreSQL: pgsql/src/backend/catalog/system_views.sql,v 1.37 2007/03/30 18:34:55 mha Exp $
  */
 
 CREATE VIEW pg_roles AS 
@@ -364,3 +364,13 @@ CREATE VIEW pg_stat_database AS
             pg_stat_get_db_tuples_updated(D.oid) AS tup_updated,
             pg_stat_get_db_tuples_deleted(D.oid) AS tup_deleted
     FROM pg_database D;
+
+CREATE VIEW pg_stat_bgwriter AS
+	SELECT
+			pg_stat_get_bgwriter_timed_checkpoints() AS checkpoints_timed,
+			pg_stat_get_bgwriter_requested_checkpoints() AS checkpoints_req,
+			pg_stat_get_bgwriter_buf_written_checkpoints() AS buffers_checkpoint,
+			pg_stat_get_bgwriter_buf_written_lru() AS buffers_lru,
+			pg_stat_get_bgwriter_buf_written_all() AS buffers_all,
+			pg_stat_get_bgwriter_maxwritten_lru() AS maxwritten_lru,
+			pg_stat_get_bgwriter_maxwritten_all() AS maxwritten_all;

src/backend/postmaster/bgwriter.c

@@ -37,7 +37,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/postmaster/bgwriter.c,v 1.36 2007/01/17 16:25:01 tgl Exp $
+ *	  $PostgreSQL: pgsql/src/backend/postmaster/bgwriter.c,v 1.37 2007/03/30 18:34:55 mha Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -50,6 +50,7 @@
 #include "access/xlog_internal.h"
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
+#include "pgstat.h"
 #include "postmaster/bgwriter.h"
 #include "storage/fd.h"
 #include "storage/freespace.h"
@@ -124,6 +125,13 @@ typedef struct
 
 static BgWriterShmemStruct *BgWriterShmem;
 
+/*
+ * BgWriter statistics counters.
+ * Stored directly in a stats message structure so it can be sent
+ * without needing to copy things around.
+ */
+PgStat_MsgBgWriter BgWriterStats;
+
 /*
  * GUC parameters
  */
@@ -242,6 +250,11 @@ BackgroundWriterMain(void)
 											 ALLOCSET_DEFAULT_MAXSIZE);
 	MemoryContextSwitchTo(bgwriter_context);
 
+	/*
+	 * Initialize statistics counters to zero
+	 */
+	memset(&BgWriterStats, 0, sizeof(BgWriterStats));
+
 	/*
 	 * If an exception is encountered, processing resumes here.
 	 *
@@ -354,6 +367,7 @@ BackgroundWriterMain(void)
 			checkpoint_requested = false;
 			do_checkpoint = true;
 			force_checkpoint = true;
+			BgWriterStats.m_requested_checkpoints++;
 		}
 		if (shutdown_requested)
 		{
@@ -376,7 +390,11 @@ BackgroundWriterMain(void)
 		now = time(NULL);
 		elapsed_secs = now - last_checkpoint_time;
 		if (elapsed_secs >= CheckPointTimeout)
+		{
 			do_checkpoint = true;
+			if (!force_checkpoint)
+				BgWriterStats.m_timed_checkpoints++;
+		}
 
 		/*
 		 * Do a checkpoint if requested, otherwise do one cycle of
@@ -473,6 +491,11 @@ BackgroundWriterMain(void)
 			}
 		}
 
+		/*
+		 * Send off activity statistics to the stats collector
+		 */
+		pgstat_send_bgwriter();
+
 		/*
 		 * Nap for the configured time, or sleep for 10 seconds if there is no
 		 * bgwriter activity configured.

src/backend/postmaster/pgstat.c

@@ -13,7 +13,7 @@
  *
  *	Copyright (c) 2001-2007, PostgreSQL Global Development Group
  *
- *	$PostgreSQL: pgsql/src/backend/postmaster/pgstat.c,v 1.151 2007/03/28 22:17:12 alvherre Exp $
+ *	$PostgreSQL: pgsql/src/backend/postmaster/pgstat.c,v 1.152 2007/03/30 18:34:55 mha Exp $
  * ----------
  */
 #include "postgres.h"
@@ -135,6 +135,18 @@ static HTAB *pgStatDBHash = NULL;
 static PgBackendStatus *localBackendStatusTable = NULL;
 static int	localNumBackends = 0;
 
+/*
+ * BgWriter global statistics counters, from bgwriter.c
+ */
+extern PgStat_MsgBgWriter BgWriterStats;
+
+/*
+ * Cluster wide statistics, kept in the stats collector.
+ * Contains statistics that are not collected per database
+ * or per table.
+ */
+static PgStat_GlobalStats globalStats;
+
 static volatile bool need_exit = false;
 static volatile bool need_statwrite = false;
 
@@ -171,6 +183,7 @@ static void pgstat_recv_resetcounter(PgStat_MsgResetcounter *msg, int len);
 static void pgstat_recv_autovac(PgStat_MsgAutovacStart *msg, int len);
 static void pgstat_recv_vacuum(PgStat_MsgVacuum *msg, int len);
 static void pgstat_recv_analyze(PgStat_MsgAnalyze *msg, int len);
+static void pgstat_recv_bgwriter(PgStat_MsgBgWriter *msg, int len);
 
 
 /* ------------------------------------------------------------
@@ -1288,6 +1301,22 @@ pgstat_fetch_stat_numbackends(void)
 	return localNumBackends;
 }
 
+/*
+ * ---------
+ * pgstat_fetch_global() -
+ *
+ *  Support function for the SQL-callable pgstat* functions. Returns
+ *  a pointer to the global statistics struct.
+ * ---------
+ */
+PgStat_GlobalStats *
+pgstat_fetch_global(void)
+{
+	backend_read_statsfile();
+
+	return &globalStats;
+}
+
 
 /* ------------------------------------------------------------
  * Functions for management of the shared-memory PgBackendStatus array
@@ -1646,6 +1675,42 @@ pgstat_send(void *msg, int len)
 #endif
 }
 
+/* ----------
+ * pgstat_send_bgwriter() -
+ *
+ *      Send bgwriter statistics to the collector
+ * ----------
+ */
+void
+pgstat_send_bgwriter(void)
+{
+	/*
+	 * This function can be called even if nothing at all has happened.
+	 * In this case, avoid sending a completely empty message to
+	 * the stats collector.
+	 */
+	if (BgWriterStats.m_timed_checkpoints == 0 &&
+		BgWriterStats.m_requested_checkpoints == 0 &&
+		BgWriterStats.m_buf_written_checkpoints == 0 &&
+		BgWriterStats.m_buf_written_lru == 0 &&
+		BgWriterStats.m_buf_written_all == 0 &&
+		BgWriterStats.m_maxwritten_lru == 0 &&
+		BgWriterStats.m_maxwritten_all == 0)
+		return;
+
+	/*
+	 * Prepare and send the message
+	 */
+	pgstat_setheader(&BgWriterStats.m_hdr, PGSTAT_MTYPE_BGWRITER);
+	pgstat_send(&BgWriterStats, sizeof(BgWriterStats));
+
+	/*
+	 * Clear out the bgwriter statistics buffer, so it can be
+	 * re-used.
+	 */
+	memset(&BgWriterStats, 0, sizeof(BgWriterStats));
+}
+
 
 /* ----------
  * PgstatCollectorMain() -
@@ -1892,6 +1957,10 @@ PgstatCollectorMain(int argc, char *argv[])
 					pgstat_recv_analyze((PgStat_MsgAnalyze *) &msg, len);
 					break;
 
+				case PGSTAT_MTYPE_BGWRITER:
+					pgstat_recv_bgwriter((PgStat_MsgBgWriter *) &msg, len);
+					break;
+
 				default:
 					break;
 			}
@@ -2030,6 +2099,11 @@ pgstat_write_statsfile(void)
 	format_id = PGSTAT_FILE_FORMAT_ID;
 	fwrite(&format_id, sizeof(format_id), 1, fpout);
 
+	/*
+	 * Write global stats struct
+	 */
+	fwrite(&globalStats, sizeof(globalStats), 1, fpout);
+
 	/*
 	 * Walk through the database table.
 	 */
@@ -2132,6 +2206,12 @@ pgstat_read_statsfile(Oid onlydb)
 	dbhash = hash_create("Databases hash", PGSTAT_DB_HASH_SIZE, &hash_ctl,
 						 HASH_ELEM | HASH_FUNCTION | HASH_CONTEXT);
 
+	/*
+	 * Clear out global statistics so they start from zero in case we can't
+	 * load an existing statsfile.
+	 */
+	memset(&globalStats, 0, sizeof(globalStats));
+
 	/*
 	 * Try to open the status file. If it doesn't exist, the backends simply
 	 * return zero for anything and the collector simply starts from scratch
@@ -2151,6 +2231,16 @@ pgstat_read_statsfile(Oid onlydb)
 		goto done;
 	}
 
+	/*
+	 * Read global stats struct
+	 */
+	if (fread(&globalStats, 1, sizeof(globalStats), fpin) != sizeof(globalStats))
+	{
+		ereport(pgStatRunningInCollector ? LOG : WARNING,
+				(errmsg("corrupted pgstat.stat file")));
+		goto done;
+	}
+
 	/*
 	 * We found an existing collector stats file. Read it and put all the
 	 * hashtable entries into place.
@@ -2656,3 +2746,22 @@ pgstat_recv_analyze(PgStat_MsgAnalyze *msg, int len)
 	tabentry->n_dead_tuples = msg->m_dead_tuples;
 	tabentry->last_anl_tuples = msg->m_live_tuples + msg->m_dead_tuples;
 }
+
+
+/* ----------
+ * pgstat_recv_bgwriter() -
+ *
+ *	Process a BGWRITER message.
+ * ----------
+ */
+static void
+pgstat_recv_bgwriter(PgStat_MsgBgWriter *msg, int len)
+{
+	globalStats.timed_checkpoints += msg->m_timed_checkpoints;
+	globalStats.requested_checkpoints += msg->m_requested_checkpoints;
+	globalStats.buf_written_checkpoints += msg->m_buf_written_checkpoints;
+	globalStats.buf_written_lru += msg->m_buf_written_lru;
+	globalStats.buf_written_all += msg->m_buf_written_all;
+	globalStats.maxwritten_lru += msg->m_maxwritten_lru;
+	globalStats.maxwritten_all += msg->m_maxwritten_all;
+}